### Run Basic Memory Setup

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/docs/getting-started.md

Initiate the setup process for Basic Memory within your project. This command starts an interview to configure project mapping, schema seeding, and skill installation.

```bash
/basic-memory:bm-setup
```

--------------------------------

### Example: Setting up an Obsidian Vault with Docker

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/Docker.md

Step-by-step guide to configure an Obsidian vault as a project within Basic Memory when using Docker volumes, including creating, setting default, and syncing.

```yaml
volumes:
  - /Users/yourname/Documents/ObsidianVault:/app/data:rw
```

```bash
# Create project pointing to mounted vault
docker exec basic-memory-server basic-memory project create obsidian /app/data
```

```bash
# Set as default
docker exec basic-memory-server basic-memory project set-default obsidian
```

```bash
# Sync to index all files
docker exec basic-memory-server basic-memory reindex
```

--------------------------------

### Setup Basic Memory CI for Cloud

Source: https://github.com/basicmachines-co/basic-memory/blob/main/src/basic_memory/ci/README.md

Installs workflow, config, prompt, and soul files for Basic Memory CI. Use --cloud flag for cloud deployment. The --yes flag bypasses confirmation prompts. Use --project-id for projects with duplicate names across workspaces.

```bash
bm cloud api-key save bmc_...
bm cloud workspace list
bm ci setup --project <project-name> --workspace <workspace-slug-or-id> --cloud --yes
```

```bash
bm ci setup --project <project-name> --project-id <project-external-id> --cloud --yes
```

--------------------------------

### Install and Run Development Commands

Source: https://github.com/basicmachines-co/basic-memory/blob/main/README.md

Use these commands for installing dependencies, running tests with different backends, performing quality checks, and managing migrations. Ensure you have 'just' installed.

```bash
just install          # Install with dev dependencies
just test-sqlite      # All tests, SQLite
just test-postgres    # All tests, Postgres (testcontainers)
just test             # Both backends
just fast-check       # fix/format/typecheck + impacted tests
just doctor           # File <-> DB consistency check (temp config)
just package-check    # Claude Code, skills, Hermes, OpenClaw package checks
just lint
just typecheck        # Pyright (primary)
just typecheck-ty     # ty (supplemental)
just format
just check            # All quality checks
just migration "msg"  # New Alembic migration
```

--------------------------------

### Install Rclone with OS Package Manager

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Install rclone using your operating system's package manager if `bm cloud setup` fails to find or install it automatically. Rerun setup after installation.

```bash
# macOS
brew install rclone
```

```bash
# Debian/Ubuntu
sudo apt install rclone
```

```bash
# Fedora
sudo dnf install rclone
```

```bash
# Arch
sudo pacman -S rclone
```

--------------------------------

### Local Development Setup for Basic Memory

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/hermes/README.md

Clone the repository and create a symbolic link to your Hermes plugins directory for local development. This avoids the need to install the plugin via `hermes plugins install`.

```bash
git clone https://github.com/basicmachines-co/basic-memory ~/code/basic-memory
mkdir -p ~/.hermes/plugins
ln -snf ~/code/basic-memory/integrations/hermes ~/.hermes/plugins/basic-memory
```

--------------------------------

### Local Plugin Setup for OpenClaw

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/DEVELOPMENT.md

Clone the repository, install dependencies, and link the local plugin for development. Ensure OpenClaw is configured to load plugins from the specified path.

```bash
git clone https://github.com/basicmachines-co/basic-memory.git
cd basic-memory/integrations/openclaw
bun install
bun run fetch-skills
openclaw plugins install -l "$PWD"
openclaw plugins enable openclaw-basic-memory --slot memory
openclaw gateway restart
```

--------------------------------

### Install Beta Version

Source: https://github.com/basicmachines-co/basic-memory/blob/main/AGENTS.md

Consumers can install the beta version of the package using pip with the `--pre` flag.

```bash
pip install basic-memory --pre
```

--------------------------------

### Rclone Setup Command

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Installs and configures rclone for use with Basic Memory cloud services. This command typically uses the system's package manager for installation.

```bash
bm cloud setup              # Install rclone via package manager and configure credentials
```

--------------------------------

### One-time setup for benchmarks

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/docs/benchmarks.md

Run these commands once to set up the benchmark environment. Ensure you are in the correct directory.

```bash
cd /path/to/basic-memory-benchmarks
just sync
```

--------------------------------

### Install Project Dependencies

Source: https://github.com/basicmachines-co/basic-memory/blob/main/CONTRIBUTING.md

Install project dependencies using either just, uv, or pip. The `.[dev]` extra installs development-specific dependencies.

```bash
# Using just (recommended)
just install
```

```bash
# Or using uv
uv install -e ".[dev]"
```

```bash
# Or using pip
pip install -e ".[dev]"
```

--------------------------------

### Two-Way Sync Local and Cloud (Initial Setup)

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

For the first-time setup of two-way synchronization, use `bm cloud bisync` with the `--resync` flag to establish a baseline between local and cloud. This is for personal workspaces only.

```bash
bm cloud bisync --name research --resync
```

--------------------------------

### Check Basic Memory Project List

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/skills/bm-setup/SKILL.md

Call this tool to confirm the Basic Memory MCP server is connected and functional. This is a prerequisite before starting the setup interview.

```bash
list_memory_projects
```

--------------------------------

### Install Dependencies with Bun

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/AGENTS.md

Installs project dependencies using Bun. Ensure Bun is installed on your system.

```bash
bun install
```

--------------------------------

### Install Development Dependencies

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/AGENTS.md

Installs the necessary development dependencies for the project.

```bash
uv sync --group dev
```

--------------------------------

### Install Manpages

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/manual-pages.md

Execute 'bm man install' to copy bundled groff pages to the ~/.local/share/man directory. This command may warn if the install root is not in your MANPATH.

```bash
bm man install        # copies bundled groff pages to ~/.local/share/man
```

--------------------------------

### Basic Memory Slash Command Examples

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/hermes/README.md

Examples demonstrating the usage of Basic Memory slash commands for searching, reading notes, checking recent updates, and capturing quick notes.

```text
/bm-search Q3 OKRs
/bm-read decisions/auth-rewrite
/bm-recent yesterday
/bm-remember Reminder: switch the staging job to the new image after the rebase lands.
```

--------------------------------

### Installing Skills via npx

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/DEVELOPMENT.md

Commands for installing all skills, a specific skill, or skills for a particular agent using the Skills CLI.

```bash
# Install all skills
npx skills add basicmachines-co/basic-memory/skills
```

```bash
# Install a specific skill
npx skills add basicmachines-co/basic-memory/skills --skill memory-tasks
```

```bash
# Install for a specific agent
npx skills add basicmachines-co/basic-memory/skills --agent claude
```

--------------------------------

### Install Basic Memory CLI with uv

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/SECURITY.md

Use this command to manually install the Basic Memory CLI using uv. This ensures the CLI is available before enabling the plugin.

```bash
uv tool install basic-memory
```

--------------------------------

### Install Basic Memory

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/semantic-search.md

Install the basic-memory package to include semantic search dependencies.

```bash
pip install basic-memory
```

--------------------------------

### Install Development Dependencies

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/README.md

Installs the necessary development dependencies using uv. Use the --extra judge flag for optional judge dependencies.

```bash
uv sync --group dev
```

```bash
uv sync --group dev --extra judge
```

--------------------------------

### SessionStart Hook - JSON Output Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/DESIGN.md

An example of the JSON output format for hook-specific output, used to pass additional context. The plugin uses plain stdout instead.

```json
{"hookSpecificOutput": {"hookEventName": "SessionStart", "additionalContext": "..."}}
```

--------------------------------

### Install Basic Memory Package

Source: https://github.com/basicmachines-co/basic-memory/blob/main/llms-install.md

Installs the Basic Memory package using recommended uv or standard pip package managers.

```bash
uv tool install basic-memory
pip install basic-memory
```

--------------------------------

### Install Basic Memory

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/skills/bm-setup/SKILL.md

Use this command to install the Basic Memory tool. Ensure you are using version 0.19.0 or later.

```bash
uv tool install basic-memory
```

```bash
pip install basic-memory
```

--------------------------------

### Write Note Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/AGENTS.md

Use write_note to capture important information. This example shows how to create a note with a title and content.

```python
write_note(title="Client Meeting - [date]", ...)
```

--------------------------------

### Example Prompts for Basic Memory

Source: https://github.com/basicmachines-co/basic-memory/blob/main/README.md

These example prompts demonstrate how to interact with Basic Memory to create notes, search for information, and retrieve project-related data. They can be used with various AI clients that support Basic Memory.

```text
"Create a note about our project architecture decisions."
```

```text
"Find information about JWT auth in my notes."
```

```text
"What have I been working on this week?"
```

--------------------------------

### Construct memory:// URLs

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/NOTE-FORMAT.md

Examples of memory:// URLs using permalinks, titles, and paths.

```text
memory://auth-approaches-2024        # By permalink
memory://Authentication Approaches   # By title (auto-resolves)
memory://project/auth-approaches     # By path
```

--------------------------------

### Install Specific Basic Memory Skill

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/README.md

Installs a particular skill, such as 'memory-tasks', from the basicmachines-co/basic-memory repository. This is useful for a more tailored setup.

```bash
npx skills add basicmachines-co/basic-memory/skills --skill memory-tasks
```

--------------------------------

### Basic Memory Configuration Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Illustrates the configuration file structure, specifying project paths, cloud routing modes, API key, and cloud host.

```json
{
  "projects": {
    "personal": "/Users/me/notes",
    "research": "/Users/me/research"
  },
  "project_modes": {
    "research": "cloud"
  },
  "cloud_api_key": "bmc_abc123...",
  "cloud_host": "https://cloud.basicmemory.com",
  "default_project": "personal"
}
```

--------------------------------

### Install Basic Memory Plugin for Codex

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/codex/README.md

Installs the Basic Memory plugin from the local repository. This is a user-level installation and requires starting a new Codex thread to load the plugin.

```bash
codex plugin marketplace add "$(git rev-parse --show-toplevel)"
codex plugin add codex@basic-memory-local
```

--------------------------------

### Set API Keys and Run Live LiteLLM Harness

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/litellm-provider.md

Export necessary API keys for providers like OpenAI and Cohere, then run the live LiteLLM harness using the 'just' command. This makes real API calls and consumes provider quota.

```bash
export OPENAI_API_KEY=sk-...
export COHERE_API_KEY=...

just test-litellm-live
```

--------------------------------

### Get Recent Activity

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/memory-continue/SKILL.md

Use `recent_activity` when there is no clear starting point to see what has changed recently and let it surface relevant threads. Specify a `timeframe` and `depth` for the search.

```python
recent_activity(timeframe="3d", depth=1)
```

--------------------------------

### Set up Basic Memory Cloud Sync

Source: https://github.com/basicmachines-co/basic-memory/blob/main/CLAUDE.md

Initiate the setup process for cloud synchronization with Basic Memory. This may involve configuration steps for connecting to the cloud.

```bash
basic-memory cloud setup
```

--------------------------------

### List and Add Projects (Local and Cloud)

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Use these commands to list existing projects and add new ones, specifying whether to target local or cloud environments.

```bash
# Local route
bm project list --local
bm project add research ~/Documents/research

# Cloud route
bm project list --cloud
bm project add research --cloud
```

--------------------------------

### Python Shim for Backwards Compatibility

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/ARCHITECTURE.md

Provides an example of a Python shim used to maintain backwards compatibility during refactoring. An old module is updated to import from a new location, and a docstring is included to guide users on the migration path.

```python
# Old module becomes a shim
from basic_memory.new_location import *

# Docstring explains migration path
"""
DEPRECATED: Import from basic_memory.new_location instead.
This shim will be removed in a future version.
"""
```

--------------------------------

### Perform Initial Project Sync (Dry Run and Actual)

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Establishes the initial sync baseline for a project. It is recommended to preview the sync with `--dry-run` before executing the actual synchronization.

```bash
# Step 1: Preview the initial sync (recommended)
bm cloud bisync --name research --resync --dry-run
```

```bash
# Step 2: If all looks good, run the actual sync
bm cloud bisync --name research --resync
```

--------------------------------

### OpenAI Provider Setup with LiteLLM

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/litellm-provider.md

Configure environment variables to use OpenAI for semantic search via LiteLLM. Ensure OPENAI_API_KEY is set.

```bash
export OPENAI_API_KEY=sk-...
export BASIC_MEMORY_SEMANTIC_SEARCH_ENABLED=true
export BASIC_MEMORY_SEMANTIC_EMBEDDING_PROVIDER=litellm
export BASIC_MEMORY_SEMANTIC_EMBEDDING_MODEL=openai/text-embedding-3-small
```

--------------------------------

### Basic Memory Cloud Setup Commands

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/BASIC_MEMORY.md

These commands are used to configure Basic Memory Cloud for optional project routing and synchronization. Authenticate with the cloud, set your API key, and then route projects to use cloud storage.

```bash
# Authenticate with Basic Memory Cloud
bm cloud login
```

```bash
# Save API key for per-project cloud routing
bm cloud set-key bmc_...
```

```bash
# Route a project through the cloud
bm project set-cloud <name>
```

```bash
# Revert a project to local routing
bm project set-local <name>
```

```bash
# Check cloud connection state
bm cloud status
```

--------------------------------

### Install MCP Python Package

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/hermes/README.md

If the `mcp` package is not auto-installed by `hermes plugins install`, use this command to install it manually. Ensure you are installing it into the Hermes virtual environment.

```bash
uv pip install --python ~/.hermes/hermes-agent/venv/bin/python mcp
```

--------------------------------

### Install Project Dependencies

Source: https://github.com/basicmachines-co/basic-memory/blob/main/AGENTS.md

Use these commands to install project dependencies. The first option installs the project in editable mode with development dependencies, while the second is a standard pip installation.

```bash
just install
```

```bash
pip install -e ".[dev]"
```

--------------------------------

### Basic Memory Configuration Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/DESIGN.md

This JSON configuration defines primary and secondary projects for recall, and team projects for sharing. It specifies a 'promoteFolder' for shared notes.

```json
{
  "basicMemory": {
    "primaryProject": "basic-memory-7020…/main",
    "secondaryProjects": ["my-team-2/main", "my-team-2/notes"],
    "teamProjects": {
      "my-team-2/notes": { "promoteFolder": "shared" }
    }
  }
}
```

--------------------------------

### NVIDIA NIM Provider Setup with LiteLLM

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/litellm-provider.md

Set up NVIDIA NIM for semantic search using LiteLLM. Includes optional custom endpoint configuration.

```bash
export NVIDIA_NIM_API_KEY=...
# Optional when using a custom or self-hosted NIM endpoint:
export NVIDIA_NIM_API_BASE=https://integrate.api.nvidia.com/v1

export BASIC_MEMORY_SEMANTIC_SEARCH_ENABLED=true
export BASIC_MEMORY_SEMANTIC_EMBEDDING_PROVIDER=litellm
export BASIC_MEMORY_SEMANTIC_EMBEDDING_MODEL=nvidia_nim/nvidia/embed-qa-4
export BASIC_MEMORY_SEMANTIC_EMBEDDING_DIMENSIONS=1024
```

--------------------------------

### Set Up Cloud Sync with Rclone

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Installs rclone, fetches tenant information, generates S3 credentials, and configures a single rclone remote named 'basic-memory-cloud'.

```bash
bm cloud setup
```

--------------------------------

### Prepare and Run Short Benchmark

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/docs/benchmarks.md

Execute the short benchmark workflow, which involves preparing the environment and then running the short benchmark. This is typically for quicker testing cycles.

```bash
just bench-prepare-short
just bench-run-short
```

--------------------------------

### One-time setup for judge metrics

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/docs/benchmarks.md

If you plan to run judge metrics, execute this command after the initial sync.

```bash
just sync-judge
```

--------------------------------

### Install Judge Extras

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/AGENTS.md

Installs additional dependencies required for the LLM-as-a-judge benchmarks.

```bash
uv sync --group dev --extra judge
```

--------------------------------

### Zod String Format: GUID Validation

Source: https://github.com/basicmachines-co/basic-memory/blob/main/src/basic_memory/mcp/ui/html/search-results-tool-ui.html

Validates strings against the GUID format.

```javascript
Sv=b("$ZodGUID",(e,t)=>{t.pattern??(t.pattern=Oh),ge.init(e,t)})
```

--------------------------------

### Configure Local Sync Path for a Project

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Set up the local synchronization path for a project that exists on the cloud but lacks a configured local path. After setup, initialize bisync with `--resync`.

```bash
bm cloud sync-setup research ~/Documents/research
bm cloud bisync --name research --resync
```

--------------------------------

### Manage Basic Memory Projects

Source: https://github.com/basicmachines-co/basic-memory/blob/main/llms-install.md

Commands for adding, listing, and setting default directories for knowledge base projects.

```bash
basic-memory project add custom-project /path/to/your/directory
basic-memory project default custom-project
basic-memory project list
```

--------------------------------

### Install Hermes Basic Memory Plugin

Source: https://github.com/basicmachines-co/basic-memory/blob/main/README.md

Install the Basic Memory plugin for Hermes. This command is used when Hermes' native plugin shape is under integrations/hermes. If subpath installs are not supported, a deprecated pointer release may be used.

```bash
hermes plugins install basicmachines-co/basic-memory --path integrations/hermes
```

--------------------------------

### Create Cloud Project with Local Sync

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Add a new project to the cloud and simultaneously set up a local directory for synchronization. This prepares the project for bisync operations.

```bash
bm project add research --cloud --local-path ~/Documents/research
```

--------------------------------

### SessionStart Hook - Parallel Queries

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/DESIGN.md

These are example queries executed in parallel by the SessionStart hook to gather context. They use structured metadata filters for deterministic recall.

```python
search_notes("", metadata_filters={"type": "task", "status": {"$in": ["active", "in-progress"]}}, project=<primary>)
```

```python
search_notes("", metadata_filters={"type": "decision", "status": "open"}, project=<primary>)
```

```python
search_notes("", metadata_filters={"type": "session"}, after_date=<recallTimeframe>, project=<primary>)
```

```python
recent_activity(timeframe=<recallTimeframe>, project=<primary>)
```

--------------------------------

### Update Installed Skills

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/README.md

Updates all installed skills to their latest versions. This ensures you are using the most recent features and bug fixes.

```bash
npx skills update
```

--------------------------------

### Just Shortcut: Run Mem0 Local

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/AGENTS.md

Runs the benchmark with the 'mem0-local' provider using a 'just' shortcut.

```bash
just bench-run-mem0-local
```

--------------------------------

### Using Memory URLs with build_context and read_note

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/memory-notes/SKILL.md

Python examples showing how to use memory:// URLs to retrieve notes and their connected context.

```python
# Get a note and its connected context
build_context(url="memory://api-design-decisions")
```

```python
# Wildcard — gather all docs
build_context(url="memory://docs/*")
```

```python
# Direct read by permalink
read_note(identifier="memory://api-design-decisions")
```

--------------------------------

### Install OpenClaw Basic Memory Plugin

Source: https://github.com/basicmachines-co/basic-memory/blob/main/README.md

Install the OpenClaw Basic Memory plugin. This plugin is package-native and published from integrations/openclaw.

```bash
openclaw plugins install @basicmemory/openclaw-basic-memory
```

--------------------------------

### Start Basic Memory Docker Container

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/Docker.md

Use this command to start the Basic Memory Docker container in detached mode.

```bash
docker-compose up -d
```

--------------------------------

### Verify Basic Memory Plugin Installation

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/README.md

Commands to verify the successful installation and status of the Basic Memory plugin within OpenClaw.

```bash
openclaw plugins list
openclaw plugins inspect openclaw-basic-memory --json
openclaw plugins doctor
```

--------------------------------

### Just Shortcut: Run Basic Memory Local

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/AGENTS.md

Runs the benchmark with the 'bm-local' provider using a 'just' shortcut.

```bash
just bench-run-bm-local
```

--------------------------------

### Read Note Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/AGENTS.md

Use read_note to retrieve the full content of a specific note. This example reads a note with the identifier 'decisions/database-choice'.

```python
read_note(identifier="decisions/database-choice")
```

--------------------------------

### Repository Structure Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/AGENTS.md

Illustrates the directory structure for Basic Memory skills, where each skill is a self-contained SKILL.md file.

```markdown
memory-tasks/SKILL.md           # Task tracking across context compaction
memory-schema/SKILL.md          # Schema lifecycle (discover, infer, validate, drift)
memory-reflect/SKILL.md         # Sleep-time memory consolidation
memory-notes/SKILL.md           # Note writing patterns and knowledge graph design
memory-defrag/SKILL.md          # Memory reorganization and cleanup
memory-metadata-search/SKILL.md # Structured metadata filtering and queries
memory-lifecycle/SKILL.md       # Entity status transitions and folder-based archival
memory-ingest/SKILL.md          # Process external input into structured entities
memory-research/SKILL.md        # Web research synthesized into Basic Memory entities
memory-literary-analysis/SKILL.md # Literary analysis knowledge graph pipeline
memory-curate/SKILL.md          # Knowledge-graph curation: orphans, relations, tags, hub notes
memory-continue/SKILL.md        # Resume prior work by rebuilding context from the graph
memory-capture/SKILL.md         # Capture a thread's current state into one coherent note
```

--------------------------------

### Search Notes Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/AGENTS.md

Use search_notes to find relevant information. This example searches for 'database decision' and limits results to 5.

```python
search_notes(query="database decision", limit=5)
```

--------------------------------

### Get Latest Benchmark Run

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/docs/benchmarks.md

Retrieves the directory of the latest benchmark run. Use this command to get the run ID for other operations.

```bash
just bench-latest-run
```

--------------------------------

### Task Note Structure Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/MEMORY_TASK_FLOW.md

Illustrates the recommended frontmatter and context structure for task notes, including status and current step.

```markdown
---
title: migrate-auth-routes
type: Task
status: active
current_step: 3
---

## Context
Auth middleware is deployed behind a flag. Monitoring error rates before full rollout.

## Plan
- [x] Add middleware
- [x] Add tests
- [x] Deploy to staging
- [ ] Validate metrics
- [ ] Roll out to production
```

--------------------------------

### Install Local Plugin

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/codex/DEVELOPMENT.md

Installs the local plugin using its marketplace identifier. This makes the plugin's skills and configurations available to Codex.

```bash
codex plugin add codex@basic-memory-local
```

--------------------------------

### Verify Basic Memory Plugin Installation

Source: https://github.com/basicmachines-co/basic-memory/blob/main/plugins/claude-code/docs/getting-started.md

Check the details of the installed Basic Memory plugin to confirm its skills and hooks are loaded correctly.

```bash
claude plugin details basic-memory@basicmachines-co
```

--------------------------------

### Defrag Plan Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/skills/memory-defrag/SKILL.md

An example of a de-fragmentation plan outlining specific changes to be made to memory files. This serves as a checklist before executing changes.

```markdown
## Defrag Plan
- [ ] Split MEMORY.md "Key People" section → memory/people.md
- [ ] Remove completed tasks older than 30 days from memory/tasks/
- [ ] Merge memory/bm-marketing-ideas.md into memory/competitive/
- [ ] Update stale project status entries in MEMORY.md
```

--------------------------------

### Azure OpenAI Provider Setup with LiteLLM

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/litellm-provider.md

Configure Azure OpenAI for semantic search through LiteLLM. Requires Azure credentials and deployment details.

```bash
export AZURE_API_KEY=...
export AZURE_API_BASE=https://<resource-name>.openai.azure.com
export AZURE_API_VERSION=2024-02-01

export BASIC_MEMORY_SEMANTIC_SEARCH_ENABLED=true
export BASIC_MEMORY_SEMANTIC_EMBEDDING_PROVIDER=litellm
export BASIC_MEMORY_SEMANTIC_EMBEDDING_MODEL=azure/<deployment-name>
export BASIC_MEMORY_SEMANTIC_EMBEDDING_DIMENSIONS=1536
```

--------------------------------

### Build Context Example

Source: https://github.com/basicmachines-co/basic-memory/blob/main/integrations/openclaw/AGENTS.md

Use build_context to explore relationships between concepts. This example explores concepts related to 'API design' up to a depth of 2.

```python
build_context(url="memory://projects/api-design", depth=2)
```

--------------------------------

### Prepare Directory for Bisync Initialization

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Add at least one file to an empty directory before running `bm cloud bisync --resync` to establish a baseline. This prevents rclone bisync errors related to empty directories.

```bash
# Create a placeholder file
echo "# Research Notes" > ~/Documents/research/README.md

# Now run bisync
bm cloud bisync --name research --resync
```

--------------------------------

### Install Basic Memory with ONNX Runtime Pin

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/semantic-search.md

For Intel Macs, install basic-memory with a specific ONNX Runtime version to ensure compatibility with FastEmbed.

```bash
pip install basic-memory 'onnxruntime<1.24'
```

--------------------------------

### Add and Sync Multiple Projects

Source: https://github.com/basicmachines-co/basic-memory/blob/main/docs/cloud-cli.md

Configure multiple projects for local sync and establish initial baselines. Daily workflow involves syncing all configured projects.

```bash
# Setup multiple projects
bm project add research --cloud --local-path ~/Documents/research
bm project add work --cloud --local-path ~/work-notes
bm project add personal --cloud --local-path ~/personal

# Establish baselines
bm cloud bisync --name research --resync
bm cloud bisync --name work --resync
bm cloud bisync --name personal --resync

# Daily workflow: sync everything
bm cloud bisync --name research
bm cloud bisync --name work
bm cloud bisync --name personal
```

--------------------------------

### Prepare Benchmark Datasets

Source: https://github.com/basicmachines-co/basic-memory/blob/main/benchmarks/docs/benchmarks.md

Synchronizes and prepares the benchmark datasets. This step is required before running benchmarks for different revisions.

```bash
cd /path/to/basic-memory-benchmarks
just sync
just bench-prepare-short
just bench-prepare-long
```